// Copyright 2009 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.

// Semaphore implementation exposed to Go.
// Intended use is provide a sleep and wakeup
// primitive that can be used in the contended case
// of other synchronization primitives.
// Thus it targets the same goal as Linux's futex,
// but it has much simpler semantics.
//
// That is, don't think of these as semaphores.
// Think of them as a way to implement sleep and wakeup
// such that every sleep is paired with a single wakeup,
// even if, due to races, the wakeup happens before the sleep.
//
// See Mullender and Cox, ``Semaphores in Plan 9,''
// http://swtch.com/semaphore.pdf

package runtime

import (
	"runtime/internal/atomic"
	"runtime/internal/sys"
	"unsafe"
)

// Asynchronous semaphore for sync.Mutex.

type semaRoot struct {
	lock  mutex
	head  *sudog
	tail  *sudog
	nwait uint32 // Number of waiters. Read w/o the lock.
}

// Prime to not correlate with any user patterns.
const semTabSize = 251

var semtable [semTabSize]struct {
	root semaRoot
	pad  [sys.CacheLineSize - unsafe.Sizeof(semaRoot{})]byte
}

//go:linkname sync_runtime_Semacquire sync.runtime_Semacquire
func sync_runtime_Semacquire(addr *uint32) {
	semacquire(addr, semaBlockProfile)
}

//go:linkname net_runtime_Semacquire net.runtime_Semacquire
func net_runtime_Semacquire(addr *uint32) {
	semacquire(addr, semaBlockProfile)
}

//go:linkname sync_runtime_Semrelease sync.runtime_Semrelease
func sync_runtime_Semrelease(addr *uint32) {
	semrelease(addr)
}

//go:linkname sync_runtime_SemacquireMutex sync.runtime_SemacquireMutex
func sync_runtime_SemacquireMutex(addr *uint32) {
	semacquire(addr, semaBlockProfile|semaMutexProfile)
}

//go:linkname net_runtime_Semrelease net.runtime_Semrelease
func net_runtime_Semrelease(addr *uint32) {
	semrelease(addr)
}

func readyWithTime(s *sudog, traceskip int) {
	if s.releasetime != 0 {
		s.releasetime = cputicks()
	}
	goready(s.g, traceskip)
}

type semaProfileFlags int

const (
	semaBlockProfile semaProfileFlags = 1 << iota
	semaMutexProfile
)

// Called from runtime.
func semacquire(addr *uint32, profile semaProfileFlags) {
	gp := getg()
	if gp != gp.m.curg {
		throw("semacquire not on the G stack")
	}

	// Easy case.
	if cansemacquire(addr) {
		return
	}

	// Harder case:
	//	increment waiter count
	//	try cansemacquire one more time, return if succeeded
	//	enqueue itself as a waiter
	//	sleep
	//	(waiter descriptor is dequeued by signaler)
	s := acquireSudog()
	root := semroot(addr)
	t0 := int64(0)
	s.releasetime = 0
	s.acquiretime = 0
	if profile&semaBlockProfile != 0 && blockprofilerate > 0 {
		t0 = cputicks()
		s.releasetime = -1
	}
	if profile&semaMutexProfile != 0 && mutexprofilerate > 0 {
		if t0 == 0 {
			t0 = cputicks()
		}
		s.acquiretime = t0
	}
	for {
		lock(&root.lock)
		// Add ourselves to nwait to disable "easy case" in semrelease.
		atomic.Xadd(&root.nwait, 1)
		// Check cansemacquire to avoid missed wakeup.
		if cansemacquire(addr) {
			atomic.Xadd(&root.nwait, -1)
			unlock(&root.lock)
			break
		}
		// Any semrelease after the cansemacquire knows we're waiting
		// (we set nwait above), so go to sleep.
		root.queue(addr, s)
		goparkunlock(&root.lock, "semacquire", traceEvGoBlockSync, 4)
		if cansemacquire(addr) {
			break
		}
	}
	if s.releasetime > 0 {
		blockevent(s.releasetime-t0, 3)
	}
	releaseSudog(s)
}

func semrelease(addr *uint32) {
	root := semroot(addr)
	atomic.Xadd(addr, 1)

	// Easy case: no waiters?
	// This check must happen after the xadd, to avoid a missed wakeup
	// (see loop in semacquire).
	if atomic.Load(&root.nwait) == 0 {
		return
	}

	// Harder case: search for a waiter and wake it.
	lock(&root.lock)
	if atomic.Load(&root.nwait) == 0 {
		// The count is already consumed by another goroutine,
		// so no need to wake up another goroutine.
		unlock(&root.lock)
		return
	}
	s := root.head
	for ; s != nil; s = s.next {
		if s.elem == unsafe.Pointer(addr) {
			atomic.Xadd(&root.nwait, -1)
			root.dequeue(s)
			break
		}
	}
	if s != nil {
		if s.acquiretime != 0 {
			t0 := cputicks()
			for x := root.head; x != nil; x = x.next {
				if x.elem == unsafe.Pointer(addr) {
					x.acquiretime = t0
					break
				}
			}
			mutexevent(t0-s.acquiretime, 3)
		}
	}
	unlock(&root.lock)
	if s != nil { // May be slow, so unlock first
		readyWithTime(s, 5)
	}
}

func semroot(addr *uint32) *semaRoot {
	return &semtable[(uintptr(unsafe.Pointer(addr))>>3)%semTabSize].root
}

func cansemacquire(addr *uint32) bool {
	for {
		v := atomic.Load(addr)
		if v == 0 {
			return false
		}
		if atomic.Cas(addr, v, v-1) {
			return true
		}
	}
}

func (root *semaRoot) queue(addr *uint32, s *sudog) {
	s.g = getg()
	s.elem = unsafe.Pointer(addr)
	s.next = nil
	s.prev = root.tail
	if root.tail != nil {
		root.tail.next = s
	} else {
		root.head = s
	}
	root.tail = s
}

func (root *semaRoot) dequeue(s *sudog) {
	if s.next != nil {
		s.next.prev = s.prev
	} else {
		root.tail = s.prev
	}
	if s.prev != nil {
		s.prev.next = s.next
	} else {
		root.head = s.next
	}
	s.elem = nil
	s.next = nil
	s.prev = nil
}

// notifyList is a ticket-based notification list used to implement sync.Cond.
//
// It must be kept in sync with the sync package.
type notifyList struct {
	// wait is the ticket number of the next waiter. It is atomically
	// incremented outside the lock.
	wait uint32

	// notify is the ticket number of the next waiter to be notified. It can
	// be read outside the lock, but is only written to with lock held.
	//
	// Both wait & notify can wrap around, and such cases will be correctly
	// handled as long as their "unwrapped" difference is bounded by 2^31.
	// For this not to be the case, we'd need to have 2^31+ goroutines
	// blocked on the same condvar, which is currently not possible.
	notify uint32

	// List of parked waiters.
	lock mutex
	head *sudog
	tail *sudog
}

// less checks if a < b, considering a & b running counts that may overflow the
// 32-bit range, and that their "unwrapped" difference is always less than 2^31.
func less(a, b uint32) bool {
	return int32(a-b) < 0
}

// notifyListAdd adds the caller to a notify list such that it can receive
// notifications. The caller must eventually call notifyListWait to wait for
// such a notification, passing the returned ticket number.
//go:linkname notifyListAdd sync.runtime_notifyListAdd
func notifyListAdd(l *notifyList) uint32 {
	// This may be called concurrently, for example, when called from
	// sync.Cond.Wait while holding a RWMutex in read mode.
	return atomic.Xadd(&l.wait, 1) - 1
}

// notifyListWait waits for a notification. If one has been sent since
// notifyListAdd was called, it returns immediately. Otherwise, it blocks.
//go:linkname notifyListWait sync.runtime_notifyListWait
func notifyListWait(l *notifyList, t uint32) {
	lock(&l.lock)

	// Return right away if this ticket has already been notified.
	if less(t, l.notify) {
		unlock(&l.lock)
		return
	}

	// Enqueue itself.
	s := acquireSudog()
	s.g = getg()
	s.ticket = t
	s.releasetime = 0
	t0 := int64(0)
	if blockprofilerate > 0 {
		t0 = cputicks()
		s.releasetime = -1
	}
	if l.tail == nil {
		l.head = s
	} else {
		l.tail.next = s
	}
	l.tail = s
	goparkunlock(&l.lock, "semacquire", traceEvGoBlockCond, 3)
	if t0 != 0 {
		blockevent(s.releasetime-t0, 2)
	}
	releaseSudog(s)
}

// notifyListNotifyAll notifies all entries in the list.
//go:linkname notifyListNotifyAll sync.runtime_notifyListNotifyAll
func notifyListNotifyAll(l *notifyList) {
	// Fast-path: if there are no new waiters since the last notification
	// we don't need to acquire the lock.
	if atomic.Load(&l.wait) == atomic.Load(&l.notify) {
		return
	}

	// Pull the list out into a local variable, waiters will be readied
	// outside the lock.
	lock(&l.lock)
	s := l.head
	l.head = nil
	l.tail = nil

	// Update the next ticket to be notified. We can set it to the current
	// value of wait because any previous waiters are already in the list
	// or will notice that they have already been notified when trying to
	// add themselves to the list.
	atomic.Store(&l.notify, atomic.Load(&l.wait))
	unlock(&l.lock)

	// Go through the local list and ready all waiters.
	for s != nil {
		next := s.next
		s.next = nil
		readyWithTime(s, 4)
		s = next
	}
}

// notifyListNotifyOne notifies one entry in the list.
//go:linkname notifyListNotifyOne sync.runtime_notifyListNotifyOne
func notifyListNotifyOne(l *notifyList) {
	// Fast-path: if there are no new waiters since the last notification
	// we don't need to acquire the lock at all.
	if atomic.Load(&l.wait) == atomic.Load(&l.notify) {
		return
	}

	lock(&l.lock)

	// Re-check under the lock if we need to do anything.
	t := l.notify
	if t == atomic.Load(&l.wait) {
		unlock(&l.lock)
		return
	}

	// Update the next notify ticket number, and try to find the G that
	// needs to be notified. If it hasn't made it to the list yet we won't
	// find it, but it won't park itself once it sees the new notify number.
	atomic.Store(&l.notify, t+1)
	for p, s := (*sudog)(nil), l.head; s != nil; p, s = s, s.next {
		if s.ticket == t {
			n := s.next
			if p != nil {
				p.next = n
			} else {
				l.head = n
			}
			if n == nil {
				l.tail = p
			}
			unlock(&l.lock)
			s.next = nil
			readyWithTime(s, 4)
			return
		}
	}
	unlock(&l.lock)
}

//go:linkname notifyListCheck sync.runtime_notifyListCheck
func notifyListCheck(sz uintptr) {
	if sz != unsafe.Sizeof(notifyList{}) {
		print("runtime: bad notifyList size - sync=", sz, " runtime=", unsafe.Sizeof(notifyList{}), "\n")
		throw("bad notifyList size")
	}
}